Recent work has proposed how socially inspired mechanisms, based on “tags” and developed within social science simulations, might be applied in peer-to-peer overlay networks to maintain high cooperation between peers even when they act selfishly. The proposed mechanism involves a dynamic re-wiring al- gorithm called “SLAC”. The algorithm assumes a random sampling service over the entire population of nodes but does not implement this itself. In this paper we re-implement SLAC on an open source peer-to-peer simulation testbed called “PEERSIM”. For the random sampling service we utilize an existing protocol called “NEWSCAST”. We present the results of some experiments we performed in which peers play the Prisoner’s Dilemma game with their neighbours. Our re- sults demonstrate that SLAC augmented with NEWSCAST produces high levels of cooperation. This increases our confidence that previous results from SLAC are generally applicable and valid and also that SLAC could have applications in real implemented systems. Finally we discuss the open issues that need to be addressed for SLAC to progress to a valuable deployable protocol.
Tag-Based Cooperation in Peer-to-Peer Networks with Newscast / A. Marcozzi; D. Hales; G. Jesi; S. Arteconi; O. Babaoglu. - STAMPA. - (2005), pp. 365-379. (Intervento presentato al convegno SOAS: International Conference on Self-Organization and Adaptation of Multi-agent and Grid Systems tenutosi a Glasgow, Scotland nel Dicembre 2005).
Tag-Based Cooperation in Peer-to-Peer Networks with Newscast
JESI, GIAN PAOLO;ARTECONI, STEFANO;BABAOGLU, OZALP
2005
Abstract
Recent work has proposed how socially inspired mechanisms, based on “tags” and developed within social science simulations, might be applied in peer-to-peer overlay networks to maintain high cooperation between peers even when they act selfishly. The proposed mechanism involves a dynamic re-wiring al- gorithm called “SLAC”. The algorithm assumes a random sampling service over the entire population of nodes but does not implement this itself. In this paper we re-implement SLAC on an open source peer-to-peer simulation testbed called “PEERSIM”. For the random sampling service we utilize an existing protocol called “NEWSCAST”. We present the results of some experiments we performed in which peers play the Prisoner’s Dilemma game with their neighbours. Our re- sults demonstrate that SLAC augmented with NEWSCAST produces high levels of cooperation. This increases our confidence that previous results from SLAC are generally applicable and valid and also that SLAC could have applications in real implemented systems. Finally we discuss the open issues that need to be addressed for SLAC to progress to a valuable deployable protocol.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
